System and method for resolving data inconsistencies with a data majority
View Patent ↗A system and method for an election and data majority mechanism that solves problems such as bit flipping, mistracking, miscaching, and I/O status errors during real-time operations. Multiple copies of data are stored on various storage media of a data processing system. Errors that occur on the storage media or on other components of the data processing system are resolved by selecting the data with the highest frequency as the data majority. The data majority is propagated throughout the storage media to correct errors.
1. A system comprising:
a read engine, wherein
said read engine is configured to be coupled to a storage device,
said storage device is configured to store a plurality of copies of data, and
said read engine is configured to read a first copy of said data, a second copy of said data, and a third copy of said data from said copies of said data;
a comparator, wherein
said comparator is coupled to said read engine, and
said comparator is configured to determine an occurrence of a data match between at least two of said first copy of data, said second copy of data, and said third copy of said data by virtue of being configured to compare said first copy of data, said second copy of data, and said third copy of said data to one another;
a majority engine, wherein
said majority engine is coupled to said comparator, and
said majority engine is configured to determine a data majority among said first copy of said data, said second copy of said data, and said third copy of said data based on said occurrence of said data match; and
a volume manager, wherein
said volume manager is configured to
if said data majority cannot be determined,
select a selected one of said copies of said data, and
disable copies of said data other than said selected one of said copies of said data.
2. The system of claim 1 , further comprising:
a storage device, wherein
said storage device is configured to store said copies of said data.
3. The system of claim 1 , further comprising:
a plurality of said storage devices, wherein
said storage devices comprise said storage device, and
each of said storage devices is configured to store a corresponding copy of said data from said copies of said data.
4. The system of claim 3 , further comprising:
a plurality of memory locations, wherein
each of said memory locations is configured to store data read from a corresponding one of said storage devices.
5. The system of claim 2 , wherein
said majority engine is further configured to determine said data majority using a highest frequency of said data match.
6. The system of claim 1 , further comprising:
a write engine, wherein
said write engine is coupled to said storage device,
said write engine is configured to maintain data consistency by virtue of being configured to overwrite a copy of said copies of said data with another copy of said copies of said data,
said copy of said copies of said data is not in said data majority, and
said another copy of said copies of said data is in said data majority.
7. The system of claim 1 , further comprising:
a write engine, wherein
said write engine is coupled to said storage device, and
said write engine is configured to store said copies of said data on said storage device by virtue of being configured to write said copies of said data to said storage device.
8. The system of claim 1 , wherein
said first copy of data, said second copy of data, and said third copy of said data are randomly selected from said copies of said data.
9. A method comprising:
resolving a data inconsistency in a data processing system, wherein
said resolving comprises determining a data majority, and
said determining said data majority comprises
reading a plurality of a plurality of copies of data stored in a storage device,
determining an occurrence of a data match by comparing said plurality of said copies of said data to one another,
if said data match has not been determined to have occurred,
reading a remaining copy of said copies of said data, and
determining said occurrence of said data match, wherein
said determining said occurrence of said data match comprises comparing said plurality of said copies of said data to one another, and
said remaining copy of said copies of said data is caused to be among said plurality of said copies of said data by said reading said remaining copy of said copies of said data, and
if said data majority cannot be determined,
selecting a selected one of said copies of said data, and
disabling copies of said data other than said selected one of said copies of said data.
10. The method of claim 9 , further comprising:
returning data from a memory location corresponding to said selected one of said copies of said data.
11. The method of claim 10 , further comprising:
returning an error condition, if said data majority cannot be determined.
12. The method of claim 10 , further comprising:
overwriting one of said copies of said data other than said selected one of said copies of said data with said selected one of said copies of said data.
13. The method of claim 10 , further comprising:
overwriting a corrupted one of said copies of said data with said data majority.
14. An apparatus comprising:
means for resolving a data inconsistency in a data processing system, wherein
said means for resolving comprises means for determining a data majority, and
said means for determining said data majority comprises
means for reading a plurality of a plurality of copies of data stored in a storage device,
means for determining an occurrence of a data match comprising means for comparing said plurality of said copies of said data to one another,
means, if said means for determining said occurrence of said data match has not determined said occurrence of said data match, for
reading a remaining copy of said copies of said data, and
determining said occurrence of said data match, wherein
said means for determining said occurrence of said data match comprises means for comparing said plurality of said copies of said data to one another, and
said remaining copy of said copies of said data is caused to be among said plurality of said copies of said data by said means for reading said remaining copy of said copies of said data, and
means, if said data majority cannot be determined, for
selecting a selected one of said copies of said data, and
disabling copies of said data other than said selected one of said copies of said data.
15. The apparatus of claim 14 , further comprising:
means for returning data from a memory location corresponding to said selected one of said copies of said data.
16. The apparatus of claim 15 , further comprising:
means for returning an error condition, if said data majority cannot be determined.
17. The apparatus of claim 15 , further comprising:
means for overwriting a corrupted one of said copies of said data with said data majority.